Ulysses,

Certainly in building, which is my experience, I expect to find the critical path running through particular parts of the project.
Typically (without constraints from design and procurement r any sectional completions) I would expect the Critical Path to start with earthworks and piling, go through the sub-structure and the building frame, then into the building envelope into the services fitting out and finally into commissioning to final hand over. If it did not follow that path, I would look at the discrepancies first to make sure the particular logic / constraints were input correctly.
I am sure other disciplines have similar rules of thumb which enables a rapid reality check.

It does not matter whether one uses manual or electronic calculation methods, the answer will be the same.

Ulysses,
In my experience there is always someone on the project who will claim to know where the critical path should be. This assumption may be based on gut feel or personal knowledge.
Either way as a planner it is something that we should hone in on, as if it is to be true the analysis should identify it to be true.
If after analysis the critical path doesn’t match what was predicted, and the person that predicted it still beleives that they were correct, you need to start the detective work. If the person who suggested the critical path isn’t happy with the end result you will quickly loose their favour.

So when considering analysis of the critical path based on someones assumption it is important to ask them lots of questions as to why they think the critical path should be where they say.
Some of the reasons behind the discrepency could be the built in logic being incorrect, the durations of some activities being incorrect, but, more often than not it will be a combination of the 2 of these things and most importantly a missing peice of information known only to the individual who is questioning the critical path.

Reasons for this missing peice of information can be as simple as the person not thinking it was important or a simple oversight on someones behalf.

It is very important that as a planner you are part magician and part detective, that you can ask the same question 5 different ways, and most importantly have faith in, and an understanding of, the answers you get out of the schedule.

If you still cant get the person to beleive the answer, give them a network chart of the schedule, a calculator & a cut lunch, explain how to calculate a critical path and let them prove you wrong.

Andy